New results on the stereoselective alkylations of malic acid derivatives supported by molecular modeling

Article abstract of DOI:10.1002/hlca.200290007

The stereoselectivity of the alkylation of dialkyl malates is dependent on steric hindrance of both ester alkyl groups. It was found that the two alkyl groups have opposite effects on diastereoselectivity. Increased steric hindrance at the C(1) carboxy group increases the anti-selectivity, whereas increased steric hindrance at the C(4) carboxy group decreases it. The results are explained by comparing the structures of the enolates, which were obtained by molecular modeling. Alkylation at C(4′) of dioxolanones, derived from benzyl-substituted malic acids, with an additional stereogenic center on the side chain is dependent on the stereogenic centers of the ring acetal and of the side chain. Alkylation at low temperatures occurs only with cis-dioxolanones having an (R)-configured side-chain stereogenic center. The corresponding trans-dioxolanone and the cis-dioxolanone with a (S)-configured side-chain stereogenic center were recovered unchanged. A rationale is presented with models of monolithiated dioxolanones obtained by ab initio calculations.

Full text of DOI:10.1002/hlca.200290007

4216
Helvetica Chimica Acta Vol. 85 (2002)
New Results on the Stereoselective Alkylations of Malic Acid Derivatives
Supported by Molecular Modeling
by Michael Sefkow*, Andreas Koch, and Erich Kleinpeter
Universit‰t Potsdam, Institut f¸r Chemie, Karl-Liebknecht-Strasse 24 25, D-14476 Golm
Dedicated to my teacher, Professor Dr. Dieter Seebach, on the occasion of his 65th birthday
The stereoselectivity of the alkylation of dialkyl malates is dependent on steric hindrance of both ester alkyl
groups. It was found that the two alkyl groups have opposite effects on diastereoselectivity. Increased steric
hindrance at the C(1) carboxy group increases the anti-selectivity, whereas increased steric hindrance at the C(4)
carboxy group decreases it. The results are explained by comparing the structures of the enolates, which were
obtained by molecular modeling. Alkylation at C(4') of dioxolanones, derived from benzyl-substituted malic
acids, with an additional stereogenic center on the side chain is dependent on the stereogenic centers of the ring
acetal and of the side chain. Alkylation at low temperatures occurs only with cis-dioxolanones having an (R)-
configured side-chain stereogenic center. The corresponding trans-dioxolanone and the cis-dioxolanone with a
(S)-configured side-chain stereogenic center were recovered unchanged. A rationale is presented with models
of monolithiated dioxolanones obtained by ab initio calculations.
1.Introduction.
a-Hydroxylated lactone lignans [1], e.g., wikstromol [2], are
interesting targets for diastereo- and enantioselective synthesis, since these natural
products possess a variety of biological activities [3]. We have recently developed a
new strategy for the enantioselective synthesis of a-hydroxylated lactone lignans [4].
The compounds were obtained in six to seven steps from malic acid (1) and in overall
yields of up to 30%. The key reactions were the two diastereoselective alkylations of
malic acid derivatives (Scheme 1), both previously described by Seebach and co-workers
[5][6].
The first stereoselective alkylation, the diastereoselective reaction of dialkyl
malates, has been frequently used in the past [7]. However, according to the original
procedure [5b] (dialkyl malate, base, À78 !À 208, then À788, electrophile, À78 ! 08,
16 h), the alkylation proceeded with average yields of ca. 50 60% and with
diastereoselectivities in the range of 9 :1 anti/syn. We have reported that the alkylation
could be carried out with improved yields when the ester, electrophile, and base were
mixed together at À788, and the mixture was warmed to 108 [4]. In the first part of this
paper, we describe the effect of differently encumbered ester groups on yield and
stereoselectivity of the benzylation of dialkyl malates.
The second stereoselective alkylation at C(2) of malic acid was achieved at the
dioxolanone stage (Scheme 1). This reaction with unsubstituted dioxolanones has been
reported [6]. We have recently shown that complete stereoselective alkylation was
achieved with dioxolanones having an (R)-configured stereogenic center on the side
chain [4].

Helvetica Chimica Acta Vol. 85 (2002)
4217
Scheme 1
We will demonstrate in the second part of this paper that alkylation of dioxolanones
with a stereogenic center at C(2) (derived from substituted malic acids) is dependent
on the configuration at the side chain (C(2)) and at the ring acetal (C(2')) (Scheme 1).
Both the steric effect of the ester groups on the alkylation of dialkyl malates and the
different reactivity of the dioxolanone stereoisomers are explained by the structures of
the corresponding intermediates (enolates or carboxylates), which were calculated by
ab initio methods.
2.Results.
2.1. Alkylation of Dialkyl Malates. We have recently examined the
dependence of both yield and diastereoselectivity on the alkylation of the symmetric
i
t
dialkyl malates 2 5 (Scheme 2; R, R' ˆ Me, Et, Pr, Bu) [4c][8]. The reactions were
carried out with benzyl bromide 6 as the electrophile and lithium hexamethyldisilazide
(LHMDS) [9] as base (Table, Entries 1 4). The major conclusions were: 1) diethyl
ester 8 and di(tert-butyl) ester 10 were obtained in high yields (91 and 94%); 2)
dimethyl ester 7 and diisopropyl ester 9 were formed in significantly lower yields (75
and 80%), although diisopropyl ester 4 was not completely converted; 3) the highest
ratio of anti/syn product was achieved with diisopropyl ester 4 (9a/9s 95 :5) and the
lowest with di(tert-butyl) ester 5 [10] (10a/10s 86 :14).
We reasoned that only one of the tert-butyl groups of malate 5 was responsible for
the low stereoselectivity, and we prepared the mixed esters, isopropyl-(tert-butyl) ester
11 and (tert-butyl)-isopropyl ester 12. The hitherto unknown esters 11 and 12 were
obtained in 85 and 20% overall yields, respectively, from malic acid (1; Scheme 3).
i
Hydrolysis of dioxolanone 13 [4a] (90% from 1) in hot PrOH in the presence of a
catalytic amount of NaHCO3 [11] afforded the ester 11 in 96% yield. The ester 12 was
prepared by esterification of the carboxy group of dioxolanone 14 ( ! 15, 76%),
followed by selective hydrolysis of the acetal moiety with 2n HCl ( ! 16, 68%), and

4218
Helvetica Chimica Acta Vol. 85 (2002)
Scheme 2
Table. Yields and Stereoselectivities of the Alkylation of Malates 2 5, 11, and 12
Entry
Educt
R
R'
Tend [8]
Products
d.r.
8 : 1
9 : 1
19 : 1
7: 1
Yield [%]
Remarks
1
2
3
4
5
6
2
3
4
5
11
12
Me
Et
iPr
tBu
iPr
tBu
Me
Et
iPr
tBu
tBu
iPr
14
14
9
9
16
16
7a/7s
8a/8s
9a/9s
10a/10s
18a/18s
19a/19s
75
91
80
(17: 1; 78%)a)
a
94(9 : 1; 90%)
90
78
)
9 : 2
4 0 : 1
a) Results of a repetition experiment (Tend ˆ 168).
esterification of the C(1) carboxy group with isourea 17 [12] ( ! 12, 42%). Alternative
routes to the diester 12 such as the base-catalyzed hydrolysis of dioxolanone 15 in
tBuOH or regioselective hydrolysis of diisopropyl ester 4 to monoester 16 followed by
re-esterification, were not successful.
The alkylations of dialkyl malates 4, 5, 11, and 12 with bromide 6 and LHMDS were
carried out simultaneously in one dewar to assure comparative results. A thin-layer
chromatogram (TLC) obtained after 14h from the reaction mixtures ( Tend ˆ 168)
(Fig. 1) already provided the most-important information, which is summarized as
follows: 1) The malates 4, 5, 11, and 12 were alkylated in good yields, and only minor
amounts of the starting esters or other by-products were detected. 2) With respect to
the stereoselectivity, the dialkyl malates can be divided into two groups: malates 4 and
12 afforded the anti-isomers 9a and 19a almost exclusively, whereas malates 5 and 11
produced compounds 10a and 18a with significantly lower diastereoselectivity.
The diastereoisomers 9a/s and 10a/s were isolated with ratios similar to those
reported previously (9a/9s 17:1; 10a/10s 9 :1). On the other hand, malates 11 and 12
were alkylated with a very different diastereoselectivity (18a/18s 4.5 :1 and 19a/19s
40 :1). It is remarkable that double benzylated dialkyl malates were obtained as minor

Helvetica Chimica Acta Vol. 85 (2002)
4219
Scheme 3
Fig. 1. TLC from the reaction mixtures of diester 4 (Lane 1), 12 (Lane 2), 11 (Lane 3), and 5 (Lane 4) after
alkylation with 1.3 equiv. of bromide 6 and 1.1 equiv. of LHMDS. Double alkylation products were observed with
4 and 12 as starting materials.
products (ca. 5%) only when malates 4 or 12 were used as starting material.
Additionally, malates with an isopropyl ester at C(1) (4 and 12) reacted slower, and the
conversion was incomplete even after a 14h reaction period. In these cases, ca. 10% of
the starting material were recovered, and 78% of the products were isolated. In
contrast, the malates with the bulkier tert-butyl ester at C(1) (5 and 11) were completely
converted after 14h, and the alkylation products were obtained in good yields ( ca.
90%) (Table 1, Entries 3 6).
2.2. Alkylation of Dioxolanones Derived from Benzyl-Substituted Malic Acids. The
(2R,2'S,4'S)-dioxolanone 20 was prepared in two steps from diisopropyl malate 9 as
described previously [4a]. The reaction also yielded the (2R,2'R,4'S)-isomer 21 and the
(2S,2'S,4'S)-isomer 22 (82%, 20/21/22 18 :1:2). Diastereoisomerically pure cis-dioxo-
lanone 20 was obtained by recrystallization from EtOH/H2O. However, this process

4220
Helvetica Chimica Acta Vol. 85 (2002)
was not reliably reproducible, and the yield for stereoisomerically pure 20 was low
(50 70%). Therefore, a mixture of the unseparated isomers was reacted with bromide
6 in presence of LHMDS at À788 (Scheme 4). Surprisingly, only one alkylation
product, dioxolanone 23, derived from dioxolanone 20, was formed according to the
1H-NMR spectrum of the crude reaction mixture. Dioxolanones 21 and 22 were
recovered unchanged. Product 23 was isolated in 69% yield from the mixture of
dioxolanones (80% yield with respect to the amount of dioxolanone 20).
Scheme 4
Other
benzyl-substituted
dioxolanones
((2R,2'S,4'S)-24/(2R,2'R,4'S)-25/
(2S,2'S,4'S)-26 6 :1:1 and (2R,2'S,4'S)-27/(2R,2'R,4'S)-28/(2S,2'S,4'S)-29 9 :2 :1)
(Scheme 5) reacted similarily [4b]. The dioxolanones 30 and 31 were obtained in
51% (68%) and 50% (67%) yield as single isomers.
The result that, from a mixture of dioxolanones, only one diastereoisomer was
alkylated was independently confirmed by alkylation experiments of stereoisomeri-
cally pure dioxolanones 20 and 21 with bromide 6 and LHMDS. Dioxolanone 20 was
alkylated within 5 h at À788 to afford product 23 whereas dioxolanone 21 was
recovered unchanged in over 90% yield under the same conditions.
3.Discussion. 3.1. Alkylation of Dialkyl Malates. The results described in Sect. 2.1
show that the diastereoselectivity of the alkylation of dialkyl malates is dependent on
steric hindrance of either of the alkyl groups. For a rationalization of this effect, we
have calculated the structures of the lithium enolates, derived from the dialkyl esters 4,

Helvetica Chimica Acta Vol. 85 (2002)
4221
Scheme 5
Fig. 2. Molecular-modeling calculations of 4 ¥ 2 Li, 12 ¥ 2 Li, 5 ¥ 2 Li, and 11 ¥ 2 Li by means of the 6-31G* split-
valence basis set
5, 11, and 12 (Fig. 2). The calculations were carried out by means of the HF/6-31G*
level of theory and were optimized without restrictions [13]. Different initial
conformations of the enolates were calculated to afford the enolate structures at their
global energy minima. These conformations were the result of a conformational search
at PM3 semiempirical level.
The minimum-energy conformations of all four enolates are of similar structure. In
these, both Li-atoms are intramolecularly coordinated providing a tricyclic enolate
structure. Interestingly, the alkoxylate Li-atom was coordinated threefold between the

4222
Helvetica Chimica Acta Vol. 85 (2002)
carbonyl O-atom at C(1), the alkoxylate O-atom, and the enolate O-atom. As a
consequence, the alkoxylate O-atom and the enolate Li-atom were positioned above
the C(2)ÀC(3)ÀC(4)ÀO plane.
Since the electrophile approaches the enolate perpendicular to the CˆC double
bond [14], steric hindrance of the alkyl groups is pivotal for either the Re- or the Si-face
attack (Fig. 3). The anti-selectivity was enhanced by a factor of two when the isopropyl
ester at C(1) (malate 4) was substituted by a tert-butyl ester (malate 12). In this case,
the Re-face was more effectively blocked because one of the Me substituents of the tBu
group is positioned exactly in the trajectory of the electrophile. On the other hand,
steric repulsion on the opposite Si-face should lower the anti-selectivity. In fact, this is
i
the case with malate 5 having a tBu group at C(4). The difference with an Pr group in
which both Me groups are positioned away from the enolate moiety is the third Me
group of the tBu moiety, which unavoidably shields the Si-face. As a result, malate 5 was
alkylated with lower stereoselectivity compared to malate 4. Consequently, the lowest
anti-selectivity was achieved with malate 11 in which the bulky tBu group is at C(4) and
i
the comparably smaller Pr group at C(1). Thus, increased steric hindrance at C(1)
increases the anti-selectivity, whereas increased steric hindrance at C(4) decreases it.
Therefore, dialkyl malate 12 may be described as a matched and dialkyl malate 11 as a
mismatched pair of differently encumbered alkyl esters.
Fig. 3. Schematic summary of the structures of 5 ¥ 2 Li, 11 ¥ 2 Li, 12 ¥ 2 Li, and 4 ¥ 2 Li (from Fig. 2) and possible
steric interactions of the alkyl ester groups with the electrophile
3.2. Alkylation of Dioxolanones. The different reactivity of dioxolanone 20 vs.
dioxolanones 21 and 22 toward alkylation may be explained by either of the following
mechanisms (Scheme 6). The initial step in both routes is deprotonation of the carboxy
group by the first equivalent of LHMDS to form the lithium carboxylates of
dioxolanones 20 22 (20 ¥ Li, 21 ¥ Li, 22 ¥ Li). This reaction is very exothermic and could
be easily monitored with an internal thermometer. The structures of 20 ¥ Li 22 ¥ Li

Helvetica Chimica Acta Vol. 85 (2002)
4223
(Fig. 4) were calculated by means of the same methods as described in Sect. 3.1. All
substituents on the aromatic ring were omitted for shorter calculation times, because
calculation of dioxolanone 20 showed that the substituents were of negligible influence.
‡
The Li -ion is coordinated in these dioxolanones to the O-atoms of the carboxy group
and to the carbonyl group of the dioxolanone moiety. The Li ¥¥¥ O distances were
between 1.92 and 2.0 ä, which is in accord with similarly coordinated Li-atoms as can
be found in several X-ray structures in the Cambridge Crystallographic Data Base (e.g.,
ref.-codes FADHEX, PAJFAH, VEYVUQ). Remarkably, the most stable conformer
of the cis-dioxolanones 20 and 22 was envelope like, whereas the trans-dioxolanone 21
shows a twist-chair conformation at the minimum energy.
Scheme 6
The mono-anions may be deprotonated by a second equivalent of LHMDS to
afford the enolates 20 ¥ 2 Li, 21 ¥ Li, 22 ¥ 2 Li (Scheme 6, Path A). The latter two are
enantiomers and have the same relative configuration. Therefore, a comparison of the
enolate structures is only necessary between 20 ¥ 2 Li and 21 ¥ 2 Li (data not shown). If
the enolates were formed in all cases, then steric interactions of the enolates with the
incoming electrophile should be responsible for the different reactivity (Scheme 6,
Path A). However, LHMDS is a bulky base and, more likely, kinetic deprotonation at
C(4') may not occur with 21 ¥ Li and 22 ¥ Li (Scheme 6, Path B).
The calculated structures of the lithium carboxylates 20 ¥ Li, 21 ¥ Li, and 22 ¥ Li
revealed that the H-atom at C(4') is shielded by either the tBu group of the dioxolanone
moiety (21 ¥ Li, dHÀC(4')¥¥¥HÀC(tert) ˆ 3.15 ä) or the CH2 group of the side chain (synaxial
interaction between HÀC(4') and HÀC(3) in 22 ¥ Li, dHÀC(4')¥¥¥HÀC(3) ˆ 2.43 ä). In the
case of dioxolanone 20 ¥ Li, the stereogenic centers do not interfere with the base and
deprotonation to the enolate is possible even at À788 (Figs. 4 and 5).
The assumption that enolization of the lithium carboxylates is selective for 20 ¥ Li at
À788 is supported by experimental evidence: upon treatment with 2.4equiv. of
LHMDS at À788 in presence of bromide 6, dioxolanone 21 was recovered unchanged
after workup in >90% yield. No alkylation or isomerization products were obtained. If
the enolate 21 ¥ 2 Li had been formed then reprotonation should occur from both
stereofaces. This was observed when the reaction of diastereoisomerically pure
dioxolanone 20 ¥ 2 Li with bromide 6 was quenched before complete consumption of
the nucleophile. In this case, product 23 was accompanied by cis-dioxolanone 20 and
the corresponding trans-dioxolanone in a 2 :1 ratio.

4224
Helvetica Chimica Acta Vol. 85 (2002)
Fig. 4. Molecular-modeling calculations of 20 ¥ Li, 21 ¥ Li, and 22 ¥ Li by means of the 6-31G* split-valence basis
set. The Li-atom is positioned in the middle of the two O-atoms at the carboxy group and the dioxolanone CˆO
group. The O ¥¥¥ Li distances are between 1.93 and 2.0 ä, which is in accord with comparable X-ray structures.
Fig. 5. Schematic summary of the structures of 20 ¥ Li, 21 ¥ Li, and 22 ¥ Li (from Fig. 4) and possible steric
interactions with the base
Conclusions.
The diastereoselectivity of the alkylation of dialkyl malates is
dependent on the steric hindrance of the employed alkyl groups. It was found that the
two alkyl groups have an opposite effect on the anti-selectivity. Increased steric
hindrance at the C(1) carboxy group increases the anti-selectivity and increased steric
hindrance at the C(4) carboxy group decreases it. Thus, the best anti-selectivity was
achieved with dialkyl malate 12 (−matched pair of alkyl esters×), and the highest amount
of syn-product was obtained with dialkyl malate 11 (−mismatched pair of alkyl esters×).
An explanation for the experimental results is given by calculated structures of the
enolates of ester 4, 5, 11, and 12. These structures revealed that a tBu group at the C(1)
carboxy group shields the Re-face most effectively whereas the Si-face is blocked when
a tBu group is at C(4).
Alkylation at C(4') of dioxolanones derived from benzyl-substituted malates was
dependent on the stereogenic centers of ring acetal and side chain. Thus, alkylation at

Helvetica Chimica Acta Vol. 85 (2002)
4225
À788 was achieved only with cis-dioxolanones having an (R)-configured stereogenic
center in the side chain (20). The corresponding trans-dioxolanone 21 and the cis-
dioxolanone 22 with (S)-configured stereogenic center were recovered unchanged.
Most probably, they were not enolized by LHMDS at À788. The structures of the
initially formed lithium carboxylates 20 ¥ Li 22 ¥ Li were calculated by ab initio
t
methods. They reveal steric repulsion of the base with either the Bu group in trans-
dioxolanone 21 ¥ Li or the vicinal CH2 group in the (S)-configured cis-dioxolanone 22 ¥
Li. This supports the assumption that 21 ¥ Li and 22 ¥ Li are not enolized at À788.
Support by the Deutsche Forschungsgemeinschaft (Se 875/1-1) and the Fonds der Chemischen Industrie is
acknowledged. We thank Mrs. A. Krtitschka for NMR and Mrs. S. F¸rstenberg for MS spectra.
Experimental Part
General. All reactions were performed in dried glassware under N2 atmosphere. Standard reagents and
solvents were purified according to known procedures [15]. Thin-layer chromatography (TLC): silica-gel plates
Merck 60 F254. Column chromatography (−flash chromatography×, FC): performed as described [16]. M.p.:
B¸chi SMP-20 apparatus, uncorrected. Optical rotations: JASCO DIP-1000 polarimeter. 1H-NMR and
13C-NMR Spectra: Bruker ARX-300; at 300.1 and 75.4MHz, resp. in CDCl 3 unless otherwise stated; d in
ppm; J in Hz. EI-MS: Finnigan MAT-SSQ-710; 70 eV; m/z (rel.).
(2S)-2-Hydroxybutanedioic Acid Di(tert-butyl) Ester (5). To an ice-cold suspension of finely powdered
malic acid (1) (2.73 g, 20.3 mmol) in CH2Cl2 (30 ml) was added N,N'-dicyclohexylcarbamimidic acid tert-butyl
ester (17; 10 ml). The suspension was warmed to r.t. when it became clear. The precipitation of urea commenced.
The reaction mixture was stirred for 7 d at r.t. Every 24h, an additional amount of 17 (2 3 ml) was added. After
7 d, the suspension was filtered through a short plug of SiO2 and washed with CH2Cl2, then with Et2O. The
solvent was removed under reduced pressure, and the residue was purified by FC (SiO2, 4 Â 30 cm; Et2O/
pentane 1:2 ! 1:0). The fraction containing the product was concentrated and further purified by bulb-to-bulb
distillation (5 Â 10À3 mbar, 1108 (air bath)) to afford 2.30 g (46%) of 5. Colorless crystals. M.p. 34.6 36.08.
1
[a]2D9 ˆ À11.7 (c ˆ 0.72, CH2Cl2). H-NMR: 3.94( td, J ˆ 5.7, 4.4, HÀC(2)); 3.20 (d, J ˆ 5.7, OH); 2.73 (dd, J ˆ
16.4, 4.4, HÀC(3)); 2.64( dd, J ˆ 16.4, 5.7, HÀC(3)); 1.49 (s, tBu); 1.46 (s, tBu). 13C-NMR: 172.77 (s); 169.76 (s);
82.62 (s); 81.32 (s); 67.55 (d); 39.89 (t); 28.05 (q); 27.95 (q). MS: 247 (100), 191 (97), 163 (57), 135 (100), 57 (74).
(2S)-2-Hydroxybutanedioic Acid 4-(tert-Butyl) 1-Isopropyl Ester (11). A suspension of (2S,4S)-2-(tert-
butyl)-5-oxo-1,3-dioxolane-4-acetic acid tert-butyl ester (13; 3.00 g, 11.6 mmol) and powdered NaHCO3 (0.30 g,
i
3.2 mmol) in PrOH (150 ml) was heated to reflux for 48 h. The solid was filtered off, and the soln. was
concentrated under reduced pressure. The residue was purified by bulb-to-bulb distillation (7.8 Â 10À1 mbar,
1108 (airbath)) to afford 2.60 g (96%) of 11. Colorless oil. [a]2D9 ˆ À8.5 (c ˆ 1.25, CH2Cl2). 1H-NMR: 5.12
(sept., J ˆ 6.3, Me2CH); 4.38 (br. q, J ˆ 5, HÀC(2)); 3.30 (br. d, J ˆ 5.1, OH); 2.75 (dd, J ˆ 16.4, 4.6, HÀC(3));
2.69 (dd, J ˆ 16.4, 5.6, HÀC(3)); 1.46 (s, tBu); 1.28 (d, J ˆ 6.3, 3 H, Me2CH); 1.27 (d, J ˆ 6.3, 3 H, Me2CH).
13C-NMR: 173.05 (s); 169.64( s); 81.38 (s); 69.60 (d); 67.37 (d); 39.70 (t); 27.97 (q); 21.66 (q); 21.63 (q). MS: 233
(100), 177 (20).
(2S,4S)-2-(tert-Butyl)-5-oxo-1,3-dioxolane-4-acetic Acid Isopropyl Ester (15). A suspension of (2S,4S)-2-
(tert-butyl)-5-oxo-1,3-dioxolane-4-acetic acid (14; 4.65 g, 23.0 mmol) in toluene (50 ml), containing three drops
of DMF, was treated with oxalyl chloride (2.15 ml, 3.18 g, 25 mmol) at 08. The solid disappeared slowly at r.t.,
and a clear soln. was obtained after ca. 4h. The soln. was decanted from some precipitate. The solvent and excess
oxalyl chloride were removed in vacuo. The remaining oily acid chloride was dissolved in CH2Cl2 (40 ml) at 08,
i
and PrOH (2.0 ml, 26.1 mmol) followed by Et3N (4.0 ml, 28.7 mmol) were added dropwise. The soln. turned
red, and a dark precipitate emerged. The mixture was stirred at r.t. for 16 h and filtered through a short plug of
Celite and SiO2. The soln. was concentrated in vacuo, and the residue was purified by FC (SiO2, 3 Â 25 cm;
AcOEt/hexane1:3) to provide 4.25 g (76%) of 15. Pale yellow oil that solidified upon standing at r.t. 1H-NMR:
5.18 (d, J ˆ 0.8, HÀC(2)); 5.07 (sept., J ˆ 6.2, Me2CH); 4.68 (ddd, J ˆ 7.6, 3.8, 0.8, HÀC(4)); 2.89 (dd, J ˆ 16.7,
3.8, 1 H, CH2CO); 2.71 (dd, J ˆ 16.7, 7.6, 1 H, CH2CO); 1.26 (d, J ˆ 6.2, Me2CH); 0.98 (s, tBu). 13C-NMR: 172.34
(s); 168.50 (s); 109.66 (d); 71.74( d); 68.84( d); 36.01 (t); 34.16 (s); 23.38 (q); 21.69 (q). MS: 245 (38), 202 (17),
187 (37), 185 (30), 145 (100).

4226
Helvetica Chimica Acta Vol. 85 (2002)
(2S)-2-Hydroxybutanedioic Acid 1-(tert-Butyl) 4-Isopropyl Ester (12). A soln. of 15 (4.20 g, 17.1 mmol) in
THF/2n HCl (50 ml/30 ml) was heated to reflux for several hours until the starting material was consumed
(monitored by TLC). (2S)-2-Hydroxybutanedioic acid 4-isopropyl ester (16) and 1 were formed. THF was
removed under reduced pressure. The aq. residue was saturated with solid NaCl and extracted with Et2O (3 Â
30 ml). (In contrast to 1, 16 is soluble in Et2O.) The org. phase was dried (MgSO4), and filtered, and the solvent
was removed in vacuo to afford 2.0 g (66%) of 16. Pale yellow oil.
The crude acid was dissolved in CH2Cl2 and 17 (2 ml) was added at r.t. The resulting suspension was stirred
for 5 d at r.t. while additional amounts of 17 (2 Â 2 ml) were added. After 5 d, the mixture was filtered through a
short plug of Celite and SiO2. The clear soln. was concentrated under reduced pressure. The residue was purified
by FC (SiO2, 3 Â 25 cm; AcOEt/hexane 1:5 ! 1:2) and bulb-to-bulb distillation (1 Â 10À1 mbar, 1308 (air bath))
providing 1.10 g (42%) of 12. Pale yellow oil. [a]2D9 ˆ À14.3 (c ˆ 1.18, CH2Cl2). 1H-NMR: 5.04( sept., J ˆ 6.2,
Me2CH); 4.35 (br. q, J ˆ 5, HÀC(2)); 3.27 (br. d, J ˆ 5.2, OH); 2.77 (dd, J ˆ 16.1, 4.8, HÀC(3)); 2.70 (dd, J ˆ
16.1, 5.9, HÀC(3)); 1.49 (s, tBu); 1.25 (d, J ˆ 6.2, 3 H, Me2CH). 13C-NMR: 172.59 (s); 169.90 (s); 82.68 (s); 68.30
(d); 67.43 (d); 39.15 (t); 27.86 (q); 21.70 (q). MS: 233 (100), 177 (74), 135 (6), 57 (12).
General Procedure 1 (GP 1): Alkylation of Dialkyl Malates with Benzyl Bromide 6. To a cold (À 768) soln.
of 1 equiv. of the malate and 1.3 equiv. of 1-(benzyloxy)-4-(bromomethyl)-2-methoxybenzene (6) in THF
(10 ml) was added 2.1 equiv. of a 1.06n soln. of lithium hexamethyldisilazide (LHMDS) in THF while the temp.
remained below À708. The mixture was warmed to ca. 108 over a period of 12 16 h. The reaction was quenched
with sat. NH4Cl (10 ml) and acidified with 2n HCl soln. to pH ꢀ 2. The aq. layer was extracted with Et2O (3 Â
5 ml). The combined org. phases were dried (MgSO4) and filtered, and the solvents were removed in vacuo. The
residue was purified by FC affording the alkylated malates as colorless oils.
(2R,3S)- and (2S,3S)-2-[4-(Benzyloxy)-3-methoxybenzyl]-3-hydroxybutanedioic Acid Dimethyl Ester (7a
and 7s). According to GP 1, with (2S)-2-hydroxybutanedioic acid dimethyl ester (2; 165 mg, 1.02 mmol), 6
(435 mg, 1.42 mmol), and LHMDS (2.0 ml, 2.12 mmol, 1.06n in THF) in THF (15 ml). FC (SiO2, 3 Â 24cm;
Et2O/pentane 1:1 ! 1:0; then 2 Â 25 cm; Et2O/CH2Cl2 1:5) afforded 300 mg (76%) (7a/7s 8 :1) of 7a and 7s.
1
Data of 7a: H-NMR: 7.43 7.23 (m, 5 arom. H); 6.81 (d, J ˆ 8.1, 1 arom. H); 6.80 (d, J ˆ 1.8, 1 arom. H);
6.72 (dd, J ˆ 8.1, 1.8, 1 arom. H); 5.09 (s, CH2O); 4.12 (dd, J ˆ 7.1, 2.5, HÀC(3)); 3.85 (s, MeO); 3.72 (s, MeO);
3.64( s, MeO); 3.36 (d, J ˆ 7.1, OH); 3.19 3.09 (m, 2 H); 2.91 (m, 1 H). MS: 388 (24), 357 (7), 237 (20), 219
(16), 177 (20), 91 (100).
Data of 7s: 1H-NMR: 6.76 (d, J ˆ 1.8, 1 arom. H); 6.64( dd, J ˆ 8.2, 1.8, 1 arom. H); 4.45 (d, J ˆ 4.7,
HÀC(3)); 3.82 (s, MeO); 3.61 (s, MeO).
(2R,3S)- and (2S,3S)-2-[4-(Benzyloxy)-3-methoxybenzyl]-3-hydroxybutanedioic Acid Diethyl Ester (8a
and 8s). According to GP 1, with (2S)-2-hydroxybutanedioic acid diethylester (3; 191 mg, 1.00 mmol), 6 (435 mg,
1.42 mmol), and LHMDS (2.0 ml, 2.12 mmol, 1.06n in THF) in THF (15 ml). FC (SiO2, 3 Â 26 cm; Et2O/
pentane 1:2 ! 1:0; then 2 Â 25 cm, Et2O/CH2Cl2 13 :87) afforded 388 mg (92%) (8a/8s 9 :1) of 8a and 8s.
1
Data of 8a: H-NMR: 7.44 7.24 (m, 5 arom. H); 6.82 (d, J ˆ 1.8, 1 arom. H); 6.81 (d, J ˆ 8.1, 1 arom. H);
6.74( dd, J ˆ 8.1, 1.8, 1 arom. H); 5.10 (s, CH2O); 4.29 4.01 (m, 2 MeCH2); 4.11 (dd, J ˆ 7.1, 2.3, HÀC(3)); 3.86
(s, MeO); 3.31 (d, J ˆ 7.1, OH); 3.16 3.09 (m, 2 H); 2.92 (dd, J ˆ 15.4, 11.1, 1 H); 1.26 (t, J ˆ 7.1, MeCH2); 1.21
(t, J ˆ 7.1, MeCH2). MS: 416 (25), 371 (9), 251 (26), 233 (22), 177 (24), 91 (100).
Data of 8s: 1H-NMR: 6.78 (d, J ˆ 1.8, 1 arom. H); 6.66 (dd, J ˆ 8.2, 1.8, 1 arom. H); 4.44 (d, J ˆ 4.6,
HÀC(3)); 3.83 (s, MeO); 3.40 (d, J ˆ 4.6, 1 H); 1.12 (t, J ˆ 7.1, MeCH2); 1.05 (t, J ˆ 7.1, MeCH2).
(2R,3S)- and (2S,3S)-2-[4-(Benzyloxy)-3-methoxybenzyl]-3-hydroxybutanedioic Acid Diisopropyl Ester
(9a and 9s). According to GP 1, with (2S)-2-hydroxybutanedioic acid diisopropyl ester (4; 220 mg, 1.01 mmol), 6
(435 mg, 1.42 mmol), and LHMDS (2.0 ml, 2.12 mmol, 1.06n in THF) in THF (15 ml). FC (SiO2, 3 Â 28 cm;
Et2O/pentane 1:2 ! 1 :0; then 2 Â 25 cm; Et2O/CH2Cl2 13 :87) afforded 360 mg (80%) of 9a and 9s (9a/9s
19 :1). Colorless oil.
Data of 9a: 1H-NMR (400 MHz): 7.44 7.27 (m, 5 arom. H); 6.82 (d, J ˆ 1.8, 1 arom. H); 6.82 (d, J ˆ 8.2,
1 arom. H); 6.74( dd, J ˆ 8.1, 1.8, 1 arom. H); 5.13 (s, CH2O); 5.08 (sept., J ˆ 6.2, Me2CH); 4.99 (sept., J ˆ 6.2,
Me2CH); 4.06 (d, J ˆ 2.8, HÀC(3)); 3.88 (s, MeO); 3.25 (br. s, OH); 3.11 (dd, J ˆ 12.4, 6.4, 1 H, CH2ÀC(2));
3.06 (ddd, J ˆ 8.1, 6.4, 2.8, HÀC(2)); 2.90 (dd, J ˆ 12.4, 8.1, 1 H, CH2ÀC(2)); 1.27 (d, J ˆ 6.2, 3 H, Me2CH); 1.25
(d, J ˆ 6.3, 3 H, Me2CH); 1.20 (d, J ˆ 6.2, 3 H, Me2CH); 1.15 (d, J ˆ 6.3, 3 H, Me2CH). MS: 444 (100), 402 (9),
385 (20), 269 (26), 207 (35), 91 (33).
Data of 9s: 1H-NMR (400 MHz): 6.77 (d, J ˆ 8.2, 1 arom. H); 6.77 (d, J ˆ 1.9, 1 arom. H); 6.66 (dd, J ˆ 8.2,
1.9, 1 arom. H); 4.38 (d, J ˆ 3.8, HÀC(3)); 1.11 (d, J ˆ 6.3, 3 H, Me2CH).
(2R,3S)- and (2S,3S)-2-[4-(Benzyloxy)-3-methoxybenzyl]-3-hydroxybutanedioic Acid Di(tert-butyl) Ester
(10a and 10s). According to GP 1, with 5 (249 mg, 1.01 mmol), 6 (435 mg, 1.42 mmol), and LHMDS (2.0 ml,

Helvetica Chimica Acta Vol. 85 (2002)
4227
2.12 mmol, 1.06n in THF) in THF (15 ml). FC (SiO2, 3 Â 30 cm; Et2O/pentane 1:2 ! 1:0; then 2 Â 25 cm, Et2O/
CH2Cl2 13 :87) affording 450 mg (94%) of 10a and 10s (10a/10s 7:1). Colorless oil.
Data of 10a: 1H-NMR: 7.44 7.24 (m, 5 arom. H); 6.81 (d, J ˆ 1.8, 1 arom. H); 6.81 (d, J ˆ 8.1, 1 arom. H);
6.73 (dd, J ˆ 8.1, 1.8, 1 arom. H); 5.11 (s, CH2O); 3.96 (dd, J ˆ 7.8, 2.8, HÀC(3)); 3.86 (s, MeO); 3.29 (d, J ˆ 7.8,
OH); 3.07 (dd, J ˆ 12.3, 6.2, 1 H, CH2ÀC(2)); 2.98 (ddd, J ˆ 8.3, 6.2, 2.8, HÀC(2)); 2.87 (dd, J ˆ 12.3, 8.3, 1 H,
CH2ÀC(2)); 1.46 (s, tBu); 1.39 (s, tBu). 13C-NMR: 172.52 (s); 171.38 (s), 149.39 (s); 146.58 (s); 137.08 (s); 131.72
(s); 128.23 (d); 127.51 (d); 127.04( d); 121.09 (d); 114.05 (d); 112.81 (d); 82.17 (s); 81.28 (s); 70.86 (t); 69.88 (d);
55.72 (q); 50.75 (d); 33.55 (t); 27.78 (q). MS: 472 (100), 360 (6), 343 (4), 91 (16), 57 (22).
Data of 10s: 1H-NMR: 6.78 (d, J ˆ 1.9, 1 arom. H); 6.78 (d, J ˆ 8.2, 1 arom. H); 6.68 (dd, J ˆ 8.2, 1.9,
1 arom. H); 4.30 (dd, J ˆ 5.4, 3.7, HÀC(3)); 3.25 (d, J ˆ 5.4, OH); 2.75 (dd, J ˆ 11.3, 3.1, 1 H, CH2ÀC(2)); 1.49
(s, tBu); 1.36 (s, tBu). 13C-NMR: 172.11 (s); 171.34( s); 149.31 (s); 146.48 (s); 137.12 (s); 132.24( s); 127.00 (d);
120.89 (d); 82.72 (s); 80.96 (s); 70.90 (t); 55.68 (q); 51.63 (d); 32.43 (t).
(2R,3S)- and (2S,3S)-2-[4-(Benzyloxy)-3-methoxybenzyl]-3-hydroxybutanedioic Acid 1-(tert-Butyl)
4-Isopropyl Ester (18a and 18s). According to GP 1, with 11 (238 mg, 1.02 mmol), 6 (396 mg, 1.29 mmol),
and LHMDS (2.0 ml, 2.12 mmol, 1.06n in THF) in THF (10 ml). FC (SiO2, 2 Â 25 cm; EtOAc/hexane 14:86;
then 2 Â 28 cm; Et2O/CH2Cl2 11 :89) afforded 423 mg (90%) of 18a and 18s (18a/18s 9 :2). Colorless oil.
1
Data of 18a: H-NMR (400 MHz): 7.44 7.27 (m, 5 arom. H); 6.81 (d, J ˆ 1.8, 1 arom. H); 6.81 (d, J ˆ 8.1,
1 arom. H); 6.74( dd, J ˆ 8.1, 1.8, 1 arom. H); 5.13 (s, CH2O); 5.08 (sept., J ˆ 6.2, Me2CH); 4.02 (d, J ˆ 2.8,
HÀC(3)); 3.88 (s, MeO); 3.08 (dd, J ˆ 13.1, 6.4, 1 H, CH2ÀC(2)); 3.00 (ddd, J ˆ 9.0, 6.4, 2.8, HÀC(2)); 2.88
(dd, J ˆ 13.1, 9.0, 1 H, CH2ÀC(2)); 1.39 (s, tBu); 1.26 (d, J ˆ 6.3, 3 H, Me2CH); 1.24( d, J ˆ 6.2, 3 H, Me2CH).
MS: 458 (44), 403 (8), 223 (12), 207 (19), 91 (100), 57 (27).
Data of 18s: 1H-NMR (400 MHz): 6.77 (d, J ˆ 8.2, 1 arom. H); 6.76 (d, J ˆ 1.9, 1 arom. H); 6.66 (dd, J ˆ 8.2,
1.9, 1 arom. H); 5.12 (s, CH2O); 5.04( sept., J ˆ 6.2, Me2CH); 4.35 (d, J ˆ 3.7, HÀC(3)); 3.87 (s, MeO); 2.82
2.75 (m, 2 H); 1.37 (s, tBu); 1.28 (d, J ˆ 6.3, 3 H, Me2CH); 1.25 (d, J ˆ 6.3, 3 H, Me2CH).
(2R,3S)- and (2S,3S)-2-[4-(Benzyloxy)-3-methoxybenzyl]-3-hydroxybutanedioic Acid 4-(tert-Butyl)
1-Isopropyl Ester (19a and 19s). According to GP 1, with 12 (237 mg, 1.02 mmol), 6 (398 mg, 1.30 mmol),
and LHMDS (2.0 ml, 2.12 mmol), 1.06n in THF) in THF (10 ml). FC (SiO2, 2 Â 25 cm; EtOAc/hexane 1:4;
then 2 Â 28 cm; Et2O/CH2Cl2 11:89) afforded 365 mg (78%) of 19a and 19s (19a/19s 40 :1). Colorless
oil.
1
Data of 19a: H-NMR (400 MHz): 7.44 7.27 (m, 5 arom. H); 6.82 (d, J ˆ 8.1, 1 arom. H); 6.81 (d, J ˆ 1.5,
1 arom. H); 6.74( dd, J ˆ 8.1, 1.5, 1 arom. H); 5.13 (s, CH2O); 5.00 (sept., J ˆ 6.2, Me2CH); 3.99 (d, J ˆ 2.6,
HÀC(3)); 3.88 (s, MeO); 3.10 (dd, J ˆ 12.9, 6.6, 1 H, CH2ÀC(2)); 3.03 (ddd, J ˆ 8.4, 6.6, 2.6, HÀC(2)); 2.89
(dd, J ˆ 12.9, 8.4, 1 H, CH2ÀC(2)); 1.47 (s, tBu); 1.21 (d, J ˆ 6.2, 3 H, Me2CH); 1.16 (d, J ˆ 6.3, 3 H, Me2CH).
MS: 458 (66), 403 (12), 343 (12), 223 (16), 207 (17), 91 (100), 57 (36).
Data of 19s: 1H-NMR: 4.31 (d, J ˆ 4.2, HÀC(3)); 1.50 (s, tBu); 1.10 (d, J ˆ 6.2, 3 H, Me2CH).
General Procedure 2 (GP 2): Alkylation of Dioxolanones. To a cold (À 788) soln. of 1 equiv. of the
dioxolanone and 1.4equiv. of the benzyl bromide in THF (10 ml) was added 2.1 equiv. of LHMDS (1.06 n in
THF) in THF (Ti <À 728). The resulting pale orange soln. was stirred for 5 h at À758 and quenched with sat.
NH4Cl soln. (10 ml). The aq. layer was acidified (pH ꢀ 2) with a 2n aq. HCl soln. and extracted with Et2O (3 Â
10 ml). The combined org. extracts were dried (MgSO4) and filtered, and the solvents were removed in vacuo.
FC of the residue provided the dioxolanones as diastereoisomerically pure compounds.
(2R,2'S,4'S)-2-{4-[4-(Benzyloxy)-3-methoxybenzyl]-2-(tert-butyl)-5-oxo-1,3-dioxolan-4-yl}-3-[4'-(benzyl-
oxy)-3'-methoxyphenyl]propanoic Acid (23). According to GP 2, with a mixture of (2R,2'S,4'S)-, (2R,2'R,4'S)-,
and (2S,2'S,4'S)-3-[4-(benzyloxy)-3-methoxyphenyl]-2-[2-(tert-butyl)-5-oxo-1,3-dioxolan-4-yl]propanoic acid
(20/21/22 18 :1:2; 0.77 g, 1.79 mmol), 6 (0.80 g, 2.60 mmol) in THF (30 ml), and LHMDS (3.7 ml, 3.9 mmol,
1.06n in THF). FC (SiO2, 3 Â 22 cm; EtOAc/hexane 1 :2) afforded 0.81 g (69%) of 23. Pale yellow foam.
1H-NMR: 9.08 (br. s, COOH); 7.45 7.25 (m, 10 arom. H); 6.81 (d, J ˆ 8.1, 1 arom. H); 6.80 (br. s, 1 arom. H);
6.78 (d, J ˆ 8.1, 1 arom. H); 6.73 (br. d, J ˆ 8.1, 1 arom. H); 6.72 (br. s, 1 arom. H); 6.67 (br. d, J ˆ 8.1,
1 arom. H); 5.13 (s, CH2O); 5.08 (s, CH2O); 4.04 (s, HÀC(2')); 3.85 (s, MeO); 3.81 (s, MeO); 3.44 (d, J ˆ 13.7,
HÀC(3)); 3.41 (br. d, J ˆ 11, HÀC(2)); 3.12 (d, J ˆ 13.7, HÀC(3)); 3.08 (br. d, J ˆ 13, 1 H, CH2ÀC(4')); 2.97
(dd, J ˆ 13, 11, 1 H, CH2ÀC(4')); 0.80 (s, tBu). 13C-NMR: 175.89 (s); 173.84( s); 149.62 (s); 149.58 (s); 147.51 (s);
147.04 (s); 137.11 (s); 136.83 (s); 130.98 (s); 128.50 (d); 128.46 (d); 127.85 (d); 127.76 (d); 127.24( d); 126.85 (s);
122.60 (d); 120.66 (d); 114.09 (d); 113.99 (d); 112.54( d); 110.04( d); 82.53 (s); 70.98 (t); 70.91 (t); 55.95 (q);
55.89 (q); 54.66 (d); 37.64( t); 34.44 (s); 32.67 (t); 23.19 (q). MS: 655 (25), 654(24), 568 (15), 317 (79), 227 (100).
(2R,2'S,4'S)-2-[4-Benzyl-2-(tert-butyl)-5-oxo-1,3-dioxolan-4-yl]-3-phenylpropanoic Acid (30). According
to GP 2, with a mixture of (2R,2'S,4'S)-, (2R,2'R,4'S)-, and (2S,2'S,4'S)-2-[2-(tert-butyl)-5-oxo-1,3-dioxolan-4-

4228
Helvetica Chimica Acta Vol. 85 (2002)
yl]-3-phenylpropanoic acid (24/25/26 6 :1:1; 1.49 g, 5.10 mmol), benzyl bromide (1.03 g, 6.00 mmol) in THF
(100 ml), and LHMDS (10 ml, 10.6 mmol, 1.06n in THF). FC (SiO2, 4 Â 20 cm, EtOAc/hexane 3 :7 ! 7:3)
afforded 1.01 g (51%) of 30: Pale yellow foam. 1H-NMR: 8.52 (br. s, COOH); 7.34 7.19 ( m, 10 arom. H); 3.93
(s, HÀC(2')); 3.60 (d, J ˆ 13.6, HÀC(3)); 3.54(br. d, J ˆ 11.3, HÀC(2)); 3.22 (d, J ˆ 13.6, HÀC(3)); 3.12
(br. d, J ˆ 12.5, 1 H, CH2ÀC(4')); 3.05 (dd, J ˆ 12.5, 11.3, 1 H, CH2ÀC(4')); 0.80 (s, tBu). 13C-NMR: 176.38 (s);
173.67 (s); 137.87 (s); 134.01 (s); 130.56 (d); 128.72 (d); 128.66 (d); 128.62 (d); 127.59 (d); 126.78 (s); 110.07 (d);
82.35 (s); 54.87 (d); 37.83 (t); 34.47 (s); 32.97 (t); 23.14( q). MS: 383 (25), 325 (2), 291 (11), 234(15), 159 (32), 91
(100), 57 (24).
(2R,2'S,4'S)-2-[2-(tert-Butyl)-4-(3-methoxybenzyl)-5-oxo-1,3-dioxolan-4-yl]-3-(3-methoxyphenyl)propanoic
Acid (31). According to GP 2, with a mixture of (2R,2'S,4'S)-, (2R,2'R,4'S)-, and (2S,2'S,4'S)-2-[2-(tert-butyl)-5-
oxo-1,3-dioxolan-4-yl]-3-(3-methoxyphenyl)propanoic acid (27/28/29 9 :2 :1; 1.10 g, 3.42 mmol), 3-methoxy-
benzyl bromide (1.00 g, 4.9 mmol) in THF (30 ml), and LHMDS (8.0 ml, 8.5 mmol, 1.06n in THF). FC (SiO2,
3 Â 20 cm; EtOAc/light petroleum 3 :7 ! 2 :1) afforded 0.77 g (50%) of 31. Pale yellow foam. 1H-NMR: 10.92
(br. s, COOH); 7.24 7.16 ( m, 2 arom. H); 6.89 6.75 (m, 6 arom. H); 4.06 (s, HÀC(2')); 3.78 (s, MeO); 3.75
(s, MeO); 3.54( d, J ˆ 13.7, HÀC(3)); 3.48 (br. d, J ˆ 12, HÀC(2)); 3.17 (d, J ˆ 13.7, HÀC(3)); 3.12 (br. d, J ˆ
12, 1 H, CH2ÀC(4')); 3.03 (t, J ˆ 12, 1 H, CH2ÀC(4')); 0.82 (s, tBu). 13C-NMR: 176.62 (s); 173.55 (s); 159.65 (s);
159.62 (s); 139.40 (s); 135.42 (s); 129.58 (d); 122.81 (d); 120.93 (d); 115.73 (d); 114.17 (d); 113.39 (s), 112.25 (d);
110.03 (d); 82.28 (s); 55.11 (q); 54.99 (q); 54.71 (d); 37.87 (t); 34.42 (s); 32.99 (t); 23.12 (q). MS: 442 (33), 311
(29), 265 (44), 235 (29), 191 (51), 189 (84), 161 (43), 121 (100), 57 (64).
Ab Initio Calculations of 4 ¥ 2 Li, 5 ¥ 2 Li, 11 ¥ 2 Li, and 12 ¥ 2 Li and of 20 ¥ Li, 21 ¥ Li, 22 ¥ Li. Total energies of
the compounds were calculated with the ab initio program GAUSSIAN 98 [13]. The calculations were carried
out at the HartreeÀFock level by means of the 6-31G* split-valence basis set. The geometry optimizations of all
conformations were performed without restrictions. The minimum-energy conformation of every compound is
shown in Figs. 2 and 4 by means of the modeling software SYBYL 6.8 [17]. The quantum-chemical calculations
were processed on SGI Octane and SGI Origin computers at Potsdam University.
REFERENCES
[1] D. C. Ayres, J. D. Loike, −Lignans. Chemical, Biological and Clinical Properties×, Cambridge University
Press, Cambridge, 1990.
[2] S. Tandon, R. P. Rastogi, Phytochemistry 1976, 15, 1789.
[3] R. S. Ward, Nat. Prod. Rep. 1999, 16, 75 and other reviews of this series.
[4] a) M. Sefkow, J. Org. Chem. 2001, 66, 2342; b) M. Sefkow, A. Kelling, U. Schilde, Tetrahedron Lett. 2001,
5101; c) M. Sefkow, Tetrahedron: Asymmetry 2001, 12, 987.
[5] a) D. Seebach, D. Wasmuth, Helv. Chim. Acta 1980, 63, 197; b) D. Seebach, J. Aebi, D. Wasmuth, Org.
Synth. Collect. Vol. VII 1990, 153.
[6] D. Seebach, R. Naef, G. Calderari, Tetrahedron 1984, 40, 1313.
¬
¬
[7] J. Gawronski, K. Gawronska, −Tartaric and Malic Acid in Synthesis×, John Wiley, New York, 1999.
[8] R. W. Dugger, J. L. Ralbovsky, D. Bryant, J. Commander, S. S. Massett, N. A. Sage, J. R. Selvidio,
Tetrahedron Lett. 1992, 33, 6763.
[9] B. H. Norman, M. L. Morris, Tetrahedron Lett. 1992, 33, 6803.
[10] C. H. Kuo, A. J. Robichaud, D. J. Rew, J. D. Bergstrom, G. D. Berger, Bioorg. Med. Chem. Lett. 1994, 4,
1591.
[11] Z.-Q. Tain, B. B. Brown, D. P. Mack, C. A. Hutton, P. A. Bartlett, J. Org. Chem. 1997, 62, 514.
[12] L. J. Mathias, Synthesis 1976, 561.
[13] Gaussian 98, Revision A.11.3, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R.
Cheeseman, V. G. Zakrzewski, J. A. Montgomery Jr., R. E. Stratmann, J. C. Burant, S. Dapprich, J. M.
Millam, A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi, B.
Mennucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson, P. Y. Ayala, Q. Cui, K.
Morokuma, N. Rega, P. Salvador, J. J. Dannenberg, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B.
Foresman, J. Cioslowski, J. V. Ortiz, A. G. Baboul, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I.
Komaromi, R. Gomperts, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara,
M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, J. L. Andres, C. Gonzalez, M. Head-
Gordon, E. S. Replogle, J. A. Pople, Gaussian, Inc., Pittsburgh PA, 2002.

Helvetica Chimica Acta Vol. 85 (2002)
4229
[14] J. E. Baldwin, L. I. Kruse, J. Chem. Soc., Chem. Commun. 1977, 233; H. O. House, W. V. Phillips, T. S. B.
Sayer, C.-C. Yau, J. Org. Chem. 1978, 43, 700.
[15] D. D. Perrin, W. L. F. Armarego, −Purification of Laboratory Chemicals×, 3rd edn., Pergamon Press, 1988.
[16] W. C. Still, M. Kahn, A. Mitra, J. Org. Chem. 1978, 43, 2923.
[17] SYBYL 6.8, Tripos Inc., 1699 South Hanley Road, St Louis, MO, 2001.
Received July 3, 2002